摘要
生成式人工智能具备设计剧本、伪造图片、合成语音等功能,现已成为电信网络诈骗分子的重要犯罪工具,由此带来了犯罪资源自动化生成、犯罪手段精确化实施等一系列的问题.侦查机关需革新侦查思路,将技术理论前瞻与犯罪治理相结合,在把握生成式人工智能技术犯罪风险的基础上,以犯罪分子所处的"使用端"视角进行深入探究,分析生成式人工智能的犯罪效能.在明确文本、图片、音频、视频四类生成式人工智能的非法运用情景后,有针对性地改进电信网络诈骗犯罪预警模型,调整案件线索、证据收集方向,建立完善的电信网络诈骗犯罪的预防、识别、控制机制.
Abstract
Generative artificial intelligence possesses functions such as designing scripts,fabricating images,and synthesizing voices.It has now become a significant criminal tool for telecommunication network fraudsters,thereby giving rise to a series of issues,including the automated generation of criminal resources and the precise implementation of criminal means.Investigation authorities are required to innovate their investigation methods,integrating forward-looking technical theories with crime governance.Based on grasping the criminal risks of generative artificial intelligence technology,they should conduct an in-depth exploration from the perspective of the"user end"where the criminals are located and analyze the criminal efficacy of generative artificial intelligence.After clarifying the illegal application scenarios of the four types of generative artificial intelligence,namely text,images,audio,and video,targeted improvements should be made to the early warning models for telecommunication network fraud crimes,adjustments should be made to the directions of the case leads and evidence collection,and a comprehensive prevention,identification,and control mechanism for telecommunication network fraud crimes should be established.
NETL
NSTL
万方数据